Merge branch 'update_avspoof' into 'master'

AVspoof: added spoof attacks to ASV protocols

Updated protocols of AVspoof database

See merge request !7

advanced/databases/avspoof/2.py

@@ -40,11 +40,19 @@ class RecognitionTraining:
         self.parameters  = parameters
 
         # verification.Database implements bob.db.verification.utils interface
-        self.db = bob.db.avspoof.verification.Database(original_directory=self.root_folder, original_extension='.wav')
+#        self.db = bob.db.avspoof.verification.Database(original_directory=self.root_folder, original_extension='.wav')
+        self.db = bob.db.avspoof.Database()
 
         # retrieve all file paths for the given parameters
-        self.objs = sorted(self.db.objects(protocol=parameters['protocol'], groups=parameters['group']), key=lambda x: x.id)
+        self.objs = sorted(self.db.objects(protocol=parameters['protocol'], groups=parameters['group'], cls=('enroll', 'probe')), key=lambda x: x.id)
+        
+#        print("protocol: %s, group: %s, objects: %d" % (parameters['protocol'], parameters['group'], len(self.objs)))
 
+        if parameters['protocol'] == "smalltest":
+            assert (len(self.objs) == 741)
+        else:
+            assert (len(self.objs) == 4973)
+        
         self.next_index       = 0
 
         self.force_start_index = force_start_index
@@ -77,7 +85,7 @@ class RecognitionTraining:
 
         # read audio data
         if self.outputs['speech'].isConnected():
-            rate, audio = scipy.io.wavfile.read(obj.make_path(self.root_folder, self.db.original_extension))
+            rate, audio = scipy.io.wavfile.read(obj.make_path(directory=self.root_folder, extension='.wav'))
             data = {
                 'value': numpy.cast['float'](audio)
             }
@@ -100,15 +108,35 @@ class RecognitionTemplates:
         self.parameters             = parameters
 
         # verification.Database implements bob.db.verification.utils interface
-        self.db                     = bob.db.avspoof.verification.Database(original_directory=self.root_folder, original_extension='.wav')
+#        self.db                     = bob.db.avspoof.verification.Database(original_directory=self.root_folder, original_extension='.wav')
+        self.db = bob.db.avspoof.Database()
 
         # template Ids are model Ids (which again are just Ids of clients) in the terminoloy of the bob.db interfaces
-        self.template_ids           = sorted(self.db.model_ids(groups=parameters['group']))
+        if parameters['protocol'] == "smalltest":
+            if parameters['group'] == "train":
+                self.template_ids = sorted([1, 3])
+            if parameters['group'] == "devel":
+                self.template_ids = sorted([15, 20])
+            if parameters['group'] == "test":
+                self.template_ids = sorted([18, 33])
+        else:
+            self.template_ids = sorted([client.id for client in self.db.clients(groups=parameters['group'])])
+        
+#        print("protocol: %s, group: %s, purpose: %s, template_ids: %d" % (parameters['protocol'], parameters['group'], parameters['purpose'], len(self.template_ids)))
+
+        if parameters['protocol'] == "smalltest":
+            assert (len(self.template_ids) == 2)
+        else:
+            if parameters['group'] == "test":
+                assert (len(self.template_ids) == 16)
+            else:
+                assert (len(self.template_ids) == 14)
+        
         self.objs = []
         # store the File objects for each template
         for template_id in self.template_ids:
-          self.objs.append(sorted(self.db.objects(protocol=parameters['protocol'], purposes=parameters['purpose'], groups=parameters['group'], model_ids = (template_id,)), key=lambda x: x.id))
-
+          self.objs.append(sorted(self.db.objects(protocol=parameters['protocol'], groups=parameters['group'], cls=parameters['purpose'], clients = (template_id,)), key=lambda x: x.id))
+        
         self.current_template_index = 0
         self.current_obj_index      = 0
         self.next_index             = 0
@@ -139,6 +167,7 @@ class RecognitionTemplates:
         # current file object that we want to read and output
         obj = self.objs[self.current_template_index][self.current_obj_index]
 
+
         if self.current_obj_index == 0:
             # we update the 'template_id' and 'client_id' only when the current template/model changes
             if self.force_start_index <= self.next_index and (self.force_end_index is None or self.force_end_index >= self.next_index):
@@ -155,7 +184,7 @@ class RecognitionTemplates:
                     self.outputs['file_id'].write({'text': obj.id}, self.next_index)
 
                 if self.outputs['speech'].isConnected():
-                    rate, audio = scipy.io.wavfile.read(obj.make_path(self.root_folder, self.db.original_extension))
+                    rate, audio = scipy.io.wavfile.read(obj.make_path(directory=self.root_folder, extension='.wav'))
                     data = {
                         'value': numpy.cast['float'](audio)
                     }
@@ -174,7 +203,7 @@ class RecognitionTemplates:
         return True
 
 
-class ProbesReal:
+class Probes:
     # Reasoning: Each client may have a number of probes. Each probe may be
     # composed of any number of sample. So, each probe receives an unique
     # identifier. Those identifiers are linked to the client identifier and
@@ -188,13 +217,75 @@ class ProbesReal:
         self.parameters  = parameters
 
         # verification.Database implements bob.db.verification.utils interface
-        self.db                     = bob.db.avspoof.verification.Database(original_directory=self.root_folder, original_extension='.wav')
+#        self.db                     = bob.db.avspoof.verification.Database(original_directory=self.root_folder, original_extension='.wav')
+        self.db = bob.db.avspoof.Database()
 
         # template Ids are model Ids (which again are just Ids of clients) in the terminoloy of the bob.db interfaces
-        self.template_ids           = sorted(self.db.model_ids(groups=parameters['group']))
-        self.objs                   = sorted(self.db.objects(protocol=parameters['protocol'],
-                                                      purposes=parameters['purpose'], groups=parameters['group']), key=lambda x: x.id)
-
+        if parameters['purpose'] == "attack":
+            self.template_ids = None
+        else:
+            if parameters['protocol'] == "smalltest":
+                if parameters['group'] == "train":
+                    self.template_ids = sorted([1, 3])
+                if parameters['group'] == "devel":
+                    self.template_ids = sorted([15, 20])
+                if parameters['group'] == "test":
+                    self.template_ids = sorted([18, 33])
+            else:
+                self.template_ids = sorted([client.id for client in self.db.clients(groups=parameters['group'])])
+        
+        self.objs                   = sorted(self.db.objects(protocol=parameters['protocol'], groups=parameters['group'], cls=parameters['purpose']), key=lambda x: x.id)
+
+#        print("protocol: %s, group: %s, purpose: %s, objects: %d" % (parameters['protocol'], parameters['group'], parameters['purpose'], len(self.objs)))
+
+        if parameters['purpose'] == "attack":
+            if parameters['protocol'] == "smalltest":
+                assert (len(self.objs) == 6910)
+            if parameters['protocol'] == "grandtest":
+                if parameters['group'] == "train":
+                    assert (len(self.objs) == 56470)
+                if parameters['group'] == "devel":
+                    assert (len(self.objs) == 56470)
+                if parameters['group'] == "test":
+                    assert (len(self.objs) == 63380)
+            if parameters['protocol'] == "physical_access":
+                if parameters['group'] == "train":
+                    assert (len(self.objs) == 38580)
+                if parameters['group'] == "devel":
+                    assert (len(self.objs) == 38580)
+                if parameters['group'] == "test":
+                    assert (len(self.objs) == 43320)
+            if parameters['protocol'] == "logical_access":
+                if parameters['group'] == "train":
+                    assert (len(self.objs) == 17890)
+                if parameters['group'] == "devel":
+                    assert (len(self.objs) == 17890)
+                if parameters['group'] == "test":
+                    assert (len(self.objs) == 20060)
+        else:
+            if parameters['protocol'] == "smalltest":
+                assert (len(self.template_ids) == 2)
+            else:
+                if parameters['group'] == "test":
+                    assert (len(self.template_ids) == 16)
+                else:
+                    assert (len(self.template_ids) == 14)
+            if parameters['protocol'] == "smalltest":
+                if parameters['group'] == "train":
+                    assert (len(self.objs) == 623)
+                if parameters['group'] == "devel":
+                    assert (len(self.objs) == 586)
+                if parameters['group'] == "test":
+                    assert (len(self.objs) == 573)
+            else:
+                if parameters['group'] == "train":
+                    assert (len(self.objs) == 4193)
+                if parameters['group'] == "devel":
+                    assert (len(self.objs) == 4225)
+                if parameters['group'] == "test":
+                    assert (len(self.objs) == 4708)
+
+        
         self.next_index  = 0
 
         self.force_start_index = force_start_index
@@ -223,21 +314,30 @@ class ProbesReal:
         if self.outputs['file_id'].isConnected():
             self.outputs['file_id'].write({'text': obj.id}, self.next_index)
 
-        if self.outputs['probe_id'].isConnected():
-            self.outputs['probe_id'].write({'text': obj.id}, self.next_index)
+        if self.template_ids is None:  # we are processing attacks
+            if self.outputs['attack_id'].isConnected():
+                self.outputs['attack_id'].write({'text': obj.id}, self.next_index)
+        else:  # we are porcessing real data
+            if self.outputs['probe_id'].isConnected():
+                self.outputs['probe_id'].write({'text': obj.id}, self.next_index)
 
         if self.outputs['client_id'].isConnected():
             self.outputs['client_id'].write({'text': obj.client_id}, self.next_index)
 
         if self.outputs['template_ids'].isConnected():
-              data = {
+            if self.template_ids is None:  # it means we are processing attacks, so template_ids is just one client_id
+                data = {
+                  'text': numpy.array([obj.client_id], dtype=numpy.string_)
+                }
+            else:
+                data = {
                   'text': numpy.array(self.template_ids, dtype=numpy.string_)
-              }
-              self.outputs['template_ids'].write(data, self.next_index)
+                }
+            self.outputs['template_ids'].write(data, self.next_index)
 
         # read audio data
         if self.outputs['speech'].isConnected():
-            rate, audio = scipy.io.wavfile.read(obj.make_path(self.root_folder, self.db.original_extension))
+            rate, audio = scipy.io.wavfile.read(obj.make_path(directory=self.root_folder, extension='.wav'))
             data = {
                 'value': numpy.cast['float'](audio)
             }
@@ -261,7 +361,46 @@ class SimpleAntispoofing:
 
         # retrieve all file paths for the given parameters
         self.objs_real = sorted(self.db.objects(protocol=parameters['protocol'], groups=parameters['group'], cls='real'), key=lambda x: x.id)
+#        print("protocol: %s, group: %s, real objects: %d" % (parameters['protocol'], parameters['group'], len(self.objs_real)))
         self.objs_attack = sorted(self.db.objects(protocol=parameters['protocol'], groups=parameters['group'], cls='attack'), key=lambda x: x.id)
+#        print("protocol: %s, group: %s, attack objects: %d" % (parameters['protocol'], parameters['group'], len(self.objs_attack)))
+
+        if parameters['protocol'] == "smalltest":
+            assert (len(self.objs_attack) == 6910)
+            if parameters['group'] == "train":
+                assert (len(self.objs_real) == 741)
+            if parameters['group'] == "devel":
+                assert (len(self.objs_real) == 695)
+            if parameters['group'] == "test":
+                assert (len(self.objs_real) == 680)
+        else:
+            if parameters['group'] == "train":
+                assert (len(self.objs_real) == 4973)
+            if parameters['group'] == "devel":
+                assert (len(self.objs_real) == 4995)
+            if parameters['group'] == "test":
+                assert (len(self.objs_real) == 5576)
+            if parameters['protocol'] == "grandtest":
+                if parameters['group'] == "train":
+                    assert (len(self.objs_attack) == 56470)
+                if parameters['group'] == "devel":
+                    assert (len(self.objs_attack) == 56470)
+                if parameters['group'] == "test":
+                    assert (len(self.objs_attack) == 63380)
+            if parameters['protocol'] == "physical_access":
+                if parameters['group'] == "train":
+                    assert (len(self.objs_attack) == 38580)
+                if parameters['group'] == "devel":
+                    assert (len(self.objs_attack) == 38580)
+                if parameters['group'] == "test":
+                    assert (len(self.objs_attack) == 43320)
+            if parameters['protocol'] == "logical_access":
+                if parameters['group'] == "train":
+                    assert (len(self.objs_attack) == 17890)
+                if parameters['group'] == "devel":
+                    assert (len(self.objs_attack) == 17890)
+                if parameters['group'] == "test":
+                    assert (len(self.objs_attack) == 20060)
 
         # we have two sets of data, so to serve them sequentially, we need to know their sizes and
         # when the second set is following after the first set, hence, this approach with offsets
@@ -330,5 +469,4 @@ class SimpleAntispoofing:
 
         self.next_index += 1
 
-        return True
-
+        return True
\ No newline at end of file